Topic: Piston failure.....again!

[scondon]

   For the third time in almost 8 years I am pulling apart an engine that has experienced piston failure. Three different engines, all with the same failure: The piston edge behind the intake valve relief cutout has broken off for whatever reason

   Last time this happened I started a thread and was content with theory that carbon was building up on the piston, and over time this reduced the valve/piston clearance until they finally started "kissing" and failure resulted. I posted some pics in that thread of the failed piston and will link it here to show the type of failure I'm talking about.

http://forums.sohc4.net/index.php?topic=84902.msg956306#msg956306

   The latest failed engine has no carbon build up on the pistons so I'm looking for other ideas about how I can avoid this in the future, if possible. Or, just hoping to hear if others have experienced this same failure so I won't feel like the engine gods are pickin' on me.

   Here are some of the variables: Piston types that failed are Wiseco 836, Cycle-X 850, and MTC 890. Megacycle 125-65 cam, Webcam 63a cam, and a long duration low lift mystery cam from the '70's.

    And some of the constants: California pump gas 89 or 91 octane. Dyna 2000 ignition, Dyna-S ignition, both usually set around 35 degrees total advance. Keihin CR 29mm carbs. Kerker exhaust with med. comp baffle. 3 different heads, but they all had +2mm intake valves(number picked from memory, I can ask Mrieck for exact size since he worked them).

   Since this last failure was on my MTC 890cc I am thinking of going with Dynoman's 915cc kit, but I can't bear the thought of spending another grand or two on an engine rebuild until I can find some way to avoid failure in 10k miles.

[scottly]

 
    And some of the constants:  Dyna 2000 ignition, Dyna-S ignition, both usually set around 35 degrees total advance. Keihin CR 29mm carbs.

These sound like the most likely suspects, IMHO. Looking at the old photo, the one piston was on the verge of turning into a puddle of molten aluminum. Lean on that cylinder, with too much total advance? Just guessing.. 

[scondon]

 
    And some of the constants:  Dyna 2000 ignition, Dyna-S ignition, both usually set around 35 degrees total advance. Keihin CR 29mm carbs.

These sound like the most likely suspects, IMHO. Looking at the old photo, the one piston was on the verge of turning into a puddle of molten aluminum. Lean on that cylinder, with too much total advance? Just guessing.. 

  Pretty sure that molten puddle was caused by me riding for near 30 miles on a blown piston

   I've tried to tune the carbs towards the rich side, and keep the ignition retarded a bit to avoid issues with pinging. Plugs have never been white(lean) when pulled, usually med. to dark brown.

    On the Dyna-S I timed the ignition to hit to the left of the two hash marks at full advance(retarded). The Dyna 2000 I set at 35 deg at full advance, sometimes going as low as 32.5 deg.(Thinking stock is somewhere around 37.5 deg. full advance).

   It could be a common mechanical issue, or a common practice that I am using, or the craptastic gas we have in California.....Or, it could be three separate issues that all culminated in the same failure. It's the not knowing that's eating me right now

[scottly]

Did the same cylinders fail? Got any pics yet? (once the piston was blown, there would be no compression, so no further meltdown? The other piston with the broken land might have failed due to the added load from "flogging the team with a dead horse"?

[Jerry Rxman Griffin aka MuthaF'er]

Sean,

You been running the #$%* outta it? I know that answer  What are you setting your ring end gaps at? Piston to cylinder clearance? Forged or cast pistons?

[scondon]

Did the same cylinders fail? Got any pics yet? (once the piston was blown, there would be no compression, so no further meltdown? The other piston with the broken land might have failed due to the added load from "flogging the team with a dead horse"?

   From what I observed on that other engine it looked like two pistons had broken lands, and that the one melted through due to friction.

   One engine failed on #1, two others failed on #2. I don't have a pic of the most recent failure, but it is a "broken land" as you see in my other pics(piston on right side of pics). Surprisingly it still had compression enough to fire since the rings are intact.

   The link I posted to was an engine cobbled together that I kinda expected to fail, so take that post with a grain of salt. Just wanted a pic that would show what I was talking about.

   I will be tearing apart the bottom ends this week to see how the big end bearings are on the rods that held the failed pistons.

[scottly]

Sean,

What are you setting your ring end gaps at?

Good point.

[RAFster122s]

Running rich
to prevent pinging is going to severely contribute to your carbon loading on the piston tops and head/valves. Your shooting yourself in the foot tuning that way, as the carbon will hold way more heat and lead to starting, running, and shutdown issues. Pinging and predetonation will be major headaches to contend with...the stuff you wont be able to hear will lead to failures long term. Knock sensors on cars are managing (with electronics) the unaudible and audible knocks/detonation/pre-detonation

[scondon]

Sean,

You been running the #$%* outta it? I know that answer  What are you setting your ring end gaps at? Piston to cylinder clearance? Forged or cast pistons?

  Oh crap, Jerry, ya HAD to go and ask me all the teknikal questions, didn't ya!?  

   I'm going to have to pull notes from the pile of years old scribbles to answer those....and I will as soon as I have a bit of time. But I can tell you this:

 I actually babied this last one a little bit because I wanted to get a few miles out of it. It's the engine I put together and did last years rally with. I don't remember ever taking it up to 9k/rpm, and stuck some awful 18/48 sprockets on the bike to keep hwy rpm's low.

  The ring end gaps on those Wiseco's in the other post were ungodly big. A bit of blowby on that one. The Cycle-X and MTC ring gaps were at spec(exact numbers are in that pile I just mentioned and will make it here soon). I'll check them again once I pull the cylinders/pistons off and see if there is anything that stands out.

[scottly]

Ring gaps that are too tight can cause broken lands; the ring expands, the ends butt together and wedge into the cylinder walls. Normally not an issue unless running NOS or boost...

[scondon]

Running rich
to prevent pinging is going to severely contribute to your carbon loading on the piston tops and head/valves. Your shooting yourself in the foot tuning that way, as the carbon will hold way more heat and lead to starting, running, and shutdown issues. Pinging and predetonation will be major headaches to contend with...the stuff you wont be able to hear will lead to failures long term. Knock sensors on cars are managing (with electronics) the unaudible and audible knocks/detonation/pre-detonation

   Thanks for that. That was what I found on my first failure; carbon loading. Although I still tend to run a tad on the rich side, I've made adjustments to my tuning habits to try and avoid that. This last failure shows no evidence of carbon build-up with the exception of the piston tops being black in color(no measurable depth of build-up).

    So far, I have had a real bear of a time trying to fully eliminate predetonation as an issue with these modified 750 engines. Jetting, needle, ignition timing, octane from 89-100 have all been tried and I haven't yet been able to make it disappear altogether. Best I have been able to do is avoid the behavior/conditions that make audible pinging happen. Typically it's a hot motor under load in 4th/5th gear trying to accelerate at 3/4+ throttle position.

[scondon]

Ring gaps that are too tight can cause broken lands; the ring expands, the ends butt together and wedge into the cylinder walls. Normally not an issue unless running NOS or boost...

   Thanks,scottly. I'm going to make a list of the possible causes that come up here, and start checking my "failure" pile. I'll be on those ring gaps tomorrow night 

[scottly]

So far, I have had a real bear of a time trying to fully eliminate predetonation as an issue with these modified 750 engines. Jetting, needle, ignition timing, octane from 89-100 have all been tried and I haven't yet been able to make it disappear altogether. Best I have been able to do is avoid the behavior/conditions that make audible pinging happen. Typically it's a hot motor under load in 4th/5th gear trying to accelerate at 3/4+ throttle position.

Bingo. Retard your total timing, and or slow down the curve.

[scondon]

So far, I have had a real bear of a time trying to fully eliminate predetonation as an issue with these modified 750 engines. Jetting, needle, ignition timing, octane from 89-100 have all been tried and I haven't yet been able to make it disappear altogether. Best I have been able to do is avoid the behavior/conditions that make audible pinging happen. Typically it's a hot motor under load in 4th/5th gear trying to accelerate at 3/4+ throttle position.

Bingo. Retard your total timing, and or slow down the curve.

  I can try going below 30 deg total advance, but that's pretty damn retarded isn't it? The curve doesn't really come into play since I'm at full advance at 3750rpm(thereabouts), and audible pinging has happened at 6k/rpm or above. With the dyna 2000 I have the option of 5 curves and have tried the one that reaches total advance at 5k/rpm, but the bike ran like absolute crap no matter how I tried to tune to it.

   I spent a lot of time a few years back trying every option available to me trying to eliminate pre-ignition entirely from these modded motors. I retarded the timing until I dared not go any further, re-jetted, moved needles, tried 4 different carb racks, put the stock airbox on, went from D8 to D9 plugs, and just about everything else I could muster with the exception of trying different needle tapers. If there's a simple answer it has completely eluded me. I have decided to just blame ethanol levels and other crap in CA gas since it's one of the only things I can't take away or change.....grrrrrrr

[scottly]

Just to eliminate a variable, were any/all of these pistons modified for the larger valves; ie larger valve reliefs fly-cut? What compression ratio's?

[TwoTired]

Sean,
What is the compression ratio?

Did you calculate the "effective compression ratio? with your cam grind?

What flame front speed are you using to calculate required timing advance?

What have you done to cool the engine cylinders commensurate with the added heat load from modifications?

Have you considered a cylinder head temperature gauge?

Cheers,

[Old Scrambler]

Just thinking and adding to the pile of considerations......Where is your piston in relation to the closing of the intake and opening of the exhaust?  You may have to advance your timing to match the cam - especially if the added duration is in the beginning and same or shorter from stock at the close.  Do you have an adjustable cam sprocket?

Another just thinking question.....is your wasted spark on the exhaust stroke causing what you hear as pinging at the higher rpms? Big cam, big intakes, big pistons......but stock exhaust valves and moderate header.....hmmmm

Oil cooler.....deeper sump pan......type of oil  ??

[754]

 After repeat failures... type of oil, was my first thought..

[scondon]

    All good stuff, guys. I really do appreciate the help in thinking this one through.

TT,
   
   I haven't "calculated" effective compression ratio. Wiseco and Cycle-X are 10.25:1 advertised, MTC's are unknown, but from the domes am guessing about the same. Heads have been milled about 0.012" for a slight bump on that. Even if I knew the exact ratio I am still stuck with two choices at the pump, 89 & 91 octane. Stock engines are running at 87?

   Flame front speed to calculate timing advance? Now yer just messing with me  

   I have oil coolers on the bikes, but they are small and I will get a larger one before going forward. Certainly these modded engines are hot running, so I make every effort to keep 'em cool. They don't "tick" when shut down, but they will get a bit of smoke out of the breather on the head when I pull off the freeway and get stuck at the lights.

    Scrambler,

    I haven't clayed the pistons when assembling to check clearances, and haven't actually measured piston/valve clearances. I assemble the top end without gaskets, and spin by hand to make sure nothing hits. I add a 0.020" copper gasket at the base or head to help ease my mind. It's not science, but it's what I do. Cams are degree'd in using a slotted sprocket, and I have set them to spec as well as retarding them a few degree's.

    Wasted spark! Now there's a thought. I don't experience any back firing, or popping, or any visible smoke out the exhaust, but this is a variable I haven't put in yet. Thanks

    Recent failed engine was running Rotella 15-40. Previous failure was running Honda HP4 20-50. First failure Castrol 20-50. Those are the three oils that I use according to availability.

[Greggo]

Subscribed, and watching closely.

[Big Jay]

Ring gaps that are too tight can cause broken lands; the ring expands, the ends butt together and wedge into the cylinder walls. Normally not an issue unless running NOS or boost...

You will see the wear markings in the cylinder wall finish if the rings butted.

[scottly]

Jay, do these pistons have off-set piston pins?

[scottly]

Sean, after thinking about it, if the rings were butted, or the valves were hitting the pistons, it would have failed on your relay hell run last year. Also, the valves would have left marks on the pistons; check out the divot at around 6:30 in the attached pic. As far as the 91 octane gas, that's what I have been running, with no noticeable pinging, and I think I might be running a higher compression ratio than you.

[754]

 How did the sides of the pistons and cylinder walls look, after the pistons failed?

[scottly]

 
    And some of the constants: California pump gas 89 or 91 octane. Dyna 2000 ignition, Dyna-S ignition, both usually set around 35 degrees total advance. Keihin CR 29mm carbs. Kerker exhaust with med. comp baffle. 3 different heads, but they all had +2mm intake valves(number picked from memory, I can ask Mrieck for exact size since he worked them).

 

Same cylinder block?